The present invention relates to network switches and switching, and more particularly, this invention relates to using a software defined network (SDN) controller to segregate and prioritize SDN-controlled routes in a switch routing table.
One goal of a SDN is to allow the network to be programmable via a SDN controller. The SDN controller is typically physically separated from any of the controlled network switches, but is not necessarily located remotely therefrom. One method that allows for programmability of the network may involve the use of the OpenFlow communication protocol. However, other applications that may allow for programmability of the network either now or in the future may be used, in addition to or in place of OpenFlow, as would be understood by one of skill in the art upon reading the present descriptions.
Other methods that allow for the network to be programmable involve more traditional approaches, such as simple network management protocol (SNMP), network configuration protocol (NetConf), etc. In future versions of OpenFlow, support may be added for programming layer 3 IPv4 and layer 3 IPv6 Forwarding Elements via OpenFlow. Layer 3 forwarding element programming via OpenFlow may add support to program the layer 3 forwarding table, also referred to as a Forwarding Information Base (FIB). In contrast to the Routing Information Base (RIB), the FIB is optimized for fast longest prefix match lookup of a destination internet protocol (IP) address and may be used for data path forwarding. OpenFlow layer 3 forwarding element programming may be used by SDN user applications to program the layer 3 forwarding tables, in some conventional uses.
However, programming the FIB does not provide the flexibility to learn routes via traditional methods and devices utilizing routing protocols, such as routing information protocol (RIP), open shortest path first (OSPF), border gateway protocol (BGP), etc., as well as via SDN utilizing a suitable application, such as OpenFlow or some other suitable application known in the art. Furthermore, programming the FIB does not provide any method to prioritize between traditional and calculated routes should there be any conflict.
These drawbacks to conventional techniques have consequences, such as layer 3 SDN programming only being capable of completely replacing current distributed routing protocols/algorithms. Accordingly, in order to overcome the drawbacks of conventional systems, it would be beneficial to have a SDN using a suitable application that is capable of complementing traditional routing protocols/algorithms.